[Mark McLeod]

Just an interesting addendum to the camera power cycling problem - the one where if it doesn't initially power up properly then you must cycle the camera power to get it going again.
We experienced a particular variation of this problem only once in Atlanta, and that was in the pits during a systems check.

For most teams a simple cycling of the robot power fixes the problem, however, if you are using the IFI backup charging circuit as is, then when the 12v source is cut off, and the backup power is off, the 7.2v battery leaves enough of a charge via feedback through the breaker panel into the RC that power to the camera is not cutoff when you throw the main breaker.
When this happens the only way to cut off power to the camera altogether is to disconnect the 7.2v battery (or try to figure out if it's legal to add a diode to block current flow back into the 12v circuits).

P.S. I edited slightly to try correcting the impression that the RC was purposely running off backup power.